Regression Techniques Research Articles

Car Price Prediction Using Machine Learning

Abstract: The objective of the "Car Price Prediction Using Machine Learning" project is to anticipate car prices based on pertinent variables by utilizing predictive modelling and advanced data analytics. The rising need for precise and active pricing systems in the automobile sector is addressed by this initiative. The system will evaluate past vehicle data, including brand, model, manufacturing year, the mileage, type of fuel, and other details, by utilizing machine learning techniques. In order to produce specific predictions, the suggested model will be trained on an extensive dataset, noticing patterns and parallels within the data. In order to create a reliable prediction model, the research focuses on using regression techniques, such as ensemble approaches or linear regression. System of measurement like as absolute mean error and R-squared coefficient will be used to evaluate the predicted accuracy of the system in order to regulate its usefulness. If this resourcefulness is employed successfully, it will have a big impact on the automobile business for clients and venders alike by giving them a tool to evaluate fair market prices and helping them make decisions. This study enhances to the body of knowledge in price analysis using machine learning and lays the footing for future improvements in the prognostication of dynamic market trends in the automobile sector.

Chemoinformatic regression methods and their applicability domain.

The growing interest in chemoinformatic model uncertainty calls for a summary of the most widely used regression techniques and how to estimate their reliability. Regression models learn a mapping from the space of explanatory variables to the space of continuous output values. Among other limitations, the predictive performance of the model is restricted by the training data used for model fitting. Identification of unusual objects by outlier detection methods can improve model performance. Additionally, proper model evaluation necessitates defining the limitations of the model, often called the applicability domain. Comparable to certain classifiers, some regression techniques come with built-in methods or augmentations to quantify their (un)certainty, while others rely on generic procedures. The theoretical background of their working principles and how to deduce specific and general definitions for their domain of applicability shall be explained.

THE INFLUENCE OF THE ATTRACTIVENESS OF PROMOTIONS AND PRODUCT DESIGN ON THE DECISION TO PURCHASE THE HONDA ALL NEW BRIO SATYA

This research aims to analyze the influence of promotional attractiveness and product design on consumer purchasing decisions. The influence of these factors on consumer behavior has become a major concern in modern marketing studies, as they have a significant role in shaping purchasing preferences. In this context, a quantitative approach is used to collect data from respondents consisting of active consumers in various industrial sectors. This research draws samples from a consumer population representing a variety of demographic backgrounds and product preferences. Survey methods are used to collect primary data related to consumer perceptions about the attractiveness of promotions and product designs as well as their purchasing decisions. Data analysis was carried out using multiple linear regression techniques to identify the relationship between the variables studied. The analysis tool uses SPSS 26. The research results show that both promotional appeal and product design have a significant influence on consumer purchasing decisions. These findings provide valuable insights for marketing practitioners to design effective strategies to increase sales and consumer satisfaction.

The U-shaped relationship between corruption and international tourism demand: A gravity model approach

The tourism industry significantly contributes to a country’s economic growth and creates a positive image of the destination. This study assesses the impact of corruption on international tourism demand. It employs two-dimensional analyses using the gravity model and a pooled ordinary least square estimator to provide a unique recognition of international tourism demand. The study utilized a conditional quantile regression technique to analyze a dataset of 200 destination countries from 1995 to 2022. The empirical results demonstrate a mixed effect of corruption on international tourism demand. The analysis reveals a non-linear relationship between corruption and international tourism demand, with the inverted-U relationship being statistically significant only at the 50th–75th quantiles and not holding at the upper and lower quantiles. The research confirms that income has a positive impact on tourism across quantiles, but the impact is disproportionate. The sub-period of 2006–2022 experienced a significant decline in gross domestic product (GDP) due to the global financial crisis and its aftershocks, which severely impacted the attractiveness of destination countries for tourists. These valuable insights can inform national tourism policies and businesses.

The impact of digital transformation on firms’ value: examining the role of ESG performance and the effect of information interaction

PurposeThis study aims to examine the impact of digital transformation on firms’ value and explore the mediating impact of ESG performance and moderating impact of information interaction.Design/methodology/approachData was collected from companies listed on the Shanghai and Shenzhen stock exchange between 2012 and 2020 with 21,488 observational samples, featuring a selection of 3,348 companies. Panel data regression techniques were used to test the mediating role of ESG performance and the moderating role of information interaction.FindingsThe study found that digital transformation can improve firms’ ESG performance, which in turn positively affects their value. The firms that engage in more interaction with outsiders benefit more from digital transformation and have a higher value.Originality/valueThis study provides new theoretical insight into improving firms’ value through digital transformation and ESG performance. It is the first to discuss and study the moderating role of information interaction in the relationship between digital transformation and firms’ value.

Advancements in Retail Price Optimization: Leveraging Machine Learning Models for Profitability and Competitiveness

Retail price optimization is essential for maximizing profitability and maintaining competitiveness in today's dynamic retail landscape. This study addresses retail price optimization as a regression problem, utilizing machine learning models to predict optimal price points for products. Leveraging factors such as product attributes, competitor pricing dynamics, and customer behaviors, regression analysis provides a structured approach to understanding the intricate relationships between variables. Among various regression techniques, the Random Forest Regressor emerges as a potent strategy, offering robustness and versatility in tackling complex pricing tasks. Results indicate that Random Forest outperforms Decision Tree and Logistic Regression models regarding accuracy, precision, recall, and overall predictive performance. With Random Forest achieving an accuracy of 94%, it demonstrates superior capability in capturing complex data patterns and making accurate predictions of retail prices. By leveraging advanced analytics and machine learning techniques, retailers can optimize pricing strategies, maximize profits, and maintain competitiveness in the market. This study underscores the importance of continuously analyzing and refining pricing strategies to gain a competitive edge in the retail landscape.

Impact of Project Planning, Monitoring And Evaluation on the Sustainability of the Millennium Village Project in Pampaida, Ikara Local Government Area of Kaduna State Nigeria

The sustainability of a project refers to the capability of a project to retain its usage over the long term. The Millennium Village Project (MVP) in Pampaida was purposely initiated to alleviate over 70% of the populace from poverty. The study examined how project planning, monitoring and evaluation influenced its sustainability, adopting the participatory and the stakeholders’ theory with the involvement of donors, government, and the community people in achieving the set goals and objectives of the Project – poverty reduction. The study applied the mixed method research technique –qualitative leading and corroborated by quantitative. We selected 368 people from the community as respondents from the 28 settlements of Pampaida to provide us with the data through structured questionnaires, interviews, focused group discussions and systematic observations. The statistical tools used in the analysis were; simple percentages, frequency tables, mean, standard deviation, and the application of the SPSS version 26 for inferential statistics while using Linear Regression Technique to test the hypothesis. In the research, hypotheses were framed in null forms to address the objectives of the study and were tested statistically, which corroborated the use of thematic analysis of interviews and observations of the research. The study found that project planning, monitoring and evaluation were significant factors that positively affect the sustainability of MVP projects. Therefore, the study recommends that; the government prioritize effective, structured and comprehensive project planning to guide the project objectives. In terms of Project execution maintenance and support, feedback mechanism and capacity-building initiative, the community needs to establish committees to implement a robust and rigorous monitoring and evaluation system.

Pathways to Sustainability in Sub-Sahara Africa: Are Institutional Quality Levels Subservient in Achieving Green GDP Growth?

AbstractThere is a growing need for a transition to green economic growth (GGDP) given that the current economic system is largely environmentally unsustainable. This study thus addresses GGDP enhancement in less developed countries using the case of Sub-Sahara Africa (SSA) countries where institutional deficiencies are typically at the root of most resource depletion and environmental degradation issues. Six institutional quality measures were used namely corruption control, government effectiveness, political stability, regulatory quality, rule of law, and voice of accountability while controlling for other factors like industrialization, energy use, and population growth in the region. The study applied a battery of second-generation panel econometric techniques in the empirical analysis after which both Bootstrap Quantile regression (BQR) technique and panel ordinary least squares (OLS) estimation approaches were applied to observe the variables' roles on GGDP advancement in SSA. From the findings, corruption control and government (policy) effectiveness favorably impact Green GDP in SSA. However, both rule of law and regulatory quality performed poorly as they were insignificant to GGDP enhancement. Furthermore, all control variables promote GGDP except for population growth. Thus, the findings buttress the need to strengthen institutions for effective governance and quality environmental regulations to enhance GGDP growth towards actualizing sustainable development goals (SDGs) in the SSA region.

Bayesian analysis of the molecular emission and dust continuum of protoplanetary disks

The MIRI instrument on board the James Webb Space Telescope probes the chemistry and dust mineralogy of the inner regions of protoplanetary disks. The observed spectra are unprecedented in their detail and reveal a rich chemistry with strong diversity between objects. This complicates interpretations that are mainly based on manual continuum subtraction and 0D slab models. We investigate the physical conditions under which the gas emits in protoplanetary disks. Based on MIRI spectra, we apply a full Bayesian analysis that provides the posterior distributions of dust and molecular properties, such as column densities and emission temperatures. To do so, we introduced the Dust Continuum Kit with Line emission from Gas (DuCKLinG), a Python-based model simultaneously describing the molecular line emission and the dust continuum of protoplanetary disks without large computational cost. The model describes the dust continuum emission by dust models with precomputed dust opacities. The molecular emission is based on LTE slab models but from extended radial ranges with gradients in column densities and emission temperatures. We compare the model to observations using Bayesian analysis with linear regression techniques to reduce the dimension of the parameter space. We benchmarked this model to a complex thermo-chemical ProDiMo model of AA\,Tau and fit the MIRI spectrum of GW\,Lup. The latter allowed for a comparison to the previous results obtained with single slab models and hand-fitted continuum. We successfully decrease the computational time of the fitting method by a factor of $80$ by eliminating linear parameters, such as the emission areas, from the Bayesian run. This approach does not significantly change the retrieved molecular parameters, and only the calculated errors on the optically thin dust masses slightly decrease. For an AA\,Tau ProDiMo mock observation, we find that the retrieved molecular conditions from DuCKLinG (column densities from $3 cm^ $ to $4 cm^ $, radial range from $0.2\ au $ to $1.2\ au $, and temperature range from about $200\ K $ to $400\ K $) fall within the true values from ProDiMo (column densities between $4 cm^ $ to $5 cm^ $, radial extent $0.1\ au $ to $6.6\ au $, and temperature range from about $120\ K $ to $1000\ K $). The smaller DuCKLinG ranges can be explained by the relative flux contributions of the different parts of ProDiMo. The parameter posterior of GW\,Lup reinforces previously found results. The previously determined column densities fall within the retrieved ranges in this study for all examined molecules ( CO2 H2O HCN and C2H2 ). Similar overlap is found for the temperatures with only the temperature range of HCN (from $570^ K$ to K$) not including the previously found value ($875\ K$). This discrepancy may be due to the simultaneous fitting of all molecules compared to the step-by-step fitting of the previous study. There is statistically significant evidence for radial temperature and column density gradients for H2O and CO2 compared to the constant temperature and column density assumed in the 0D slab models. Additionally HCN and C2H2 emit from a small region with near constant conditions. Due to the small selected wavelength range $13.6\ m$ to $16.3\ m$, the dust properties are not well constrained for GW\,Lup. DuCKLinG can become an important tool to analyse the molecular emission and dust mineralogy of large samples based on JWST/MIRI spectra in an automated way.

Distributed Heterogeneity Learning for Generalized Partially Linear Models with Spatially Varying Coefficients1

Spatial heterogeneity is of great importance in social, economic, and environmental science studies. The spatially varying coefficient model is a popular and effective spatial regression technique to address spatial heterogeneity. However, accounting for heterogeneity comes at the cost of reducing model parsimony. To balance flexibility and parsimony, this article develops a class of generalized partially linear spatially varying coefficient models which allow the inclusion of both constant and spatially varying effects of covariates. Another significant challenge in many applications comes from the enormous size of the spatial datasets collected from modern technologies. To tackle this challenge, we design a novel distributed heterogeneity learning (DHL) method based on bivariate spline smoothing over a triangulation of the domain. The proposed DHL algorithm has a simple, scalable, and communication-efficient implementation scheme that can almost achieve linear speedup. In addition, this article provides rigorous theoretical support for the DHL framework. We prove that the DHL constant coefficient estimators are asymptotic normal and the DHL spline estimators reach the same convergence rate as the global spline estimators obtained using the entire dataset. The proposed DHL method is evaluated through extensive simulation studies and analyses of U.S. loan application data.

Predictive models for treated clayey soils using waste powdered glass and expanded polystyrene beads using regression analysis and artificial neural network

AbstractWaste materials contribute to a wide range of environmental and economic problems. To minimize their effects, a safe strategy for reducing such negative impact is required. Recycling and reusing waste materials have proved to be effective measures in this respect. In this study, an eco-friendly treatment is investigated based on using waste powdered glass (WGP) and EPS beads (EPSb) as mechanical and chemical admixers in soils. For this purpose, Atterberg limit, standard proctor, free swell, and unconfined compression tests are performed on soil samples with different ratios of waste materials at their optimum moisture contents. The obtained test results indicate that adding WGP to cohesive soils increases the unconfined compressive strength (UCS) and reduces free swell (FS). In contrast, using EPSb reduces both FS and UCS of the treated soil samples. An optimum combination of both waste materials is determined for the improvement of the properties of high plasticity clay used in this study. Furthermore, multiple linear regression (MLR) and artificial neural network (ANN) methods are used to predict the FS and UCS of the clayey soils based on the data obtained here and the experimental test results reported in the literature. Once the FS and UCS values of untreated soil and additive percentages are defined as independent variables, both methods are shown to predict the FS and UCS values of the treated soil samples on a satisfactory level with the coefficient of correlation ($${R}^{2}$$ R 2 ) values greater than 0.926. Additionally, when only the index properties (liquid limit, plastic limit, and plasticity index) of the soil samples with waste materials are used as dependent variables, the $${R}^{2}$$ R 2 values obtained by the ANN method are 0.968 and 0.974 for FS and UCS, respectively. The results of the untreated soil samples' FS and UCS tests are known, and the linear regression and ANN techniques yield similar results. Lastly, the ANN method is used to predict the FS and UCS of the treated samples in accordance to the limited predictors (e.g., only the Atterberg limits of the soil sample).

Do Labour Welfare Policies Matter for Workers? Evidence from the Garment Supply Chain Industry in Bangladesh

AbstractThe paper focusses on workers’ opinions in regard to welfare policies enacted under the Bangladesh Labour Act 2006. The study investigates two questions: Do labour welfare policies under labour law really matter for workers? If so, to what extent are such welfare policies practised in the global garment supply chain industry in Bangladesh? The Binary Logistic Regression technique has been used to assess the opinions of garment workers as to whether their level of satisfaction increases by workers’ welfare facility interventions. The data were collected from compliant and export-oriented factories between July 2021 and August 2021 in three Upazilas-Savar Ashulia (Dhaka District) and Gazipur Sadar (Gazipur District) in Bangladesh. This study demonstrates that overtime facilities, medical facilities, safety measures, restroom facilities, lunchroom facilities, and childcare facilities can play a pivotal role in increasing the satisfaction level of the workers. Given the similar socio-economic and geographical backgrounds, this research can be applied to the garment manufacturing industries of other South Asian countries. It serves as a valuable tool for understanding workers' attitudes and beliefs about welfare provisions. By applying the findings of this study, one can better comprehend the needs of the workers in terms of various labor welfare motivational tools.

Non-financial incentives and performance of teaching hospitals in Enugu State Nigeria

This study examined non-financial incentives and performance of teaching hospitals in Enugu State.. The study adopted survey design. The population consists of junior, senior and management staff of both teaching hospitals, numbering 1813, out of which 328 were sampled for the study. A structured questionnaire were used for data collection. Simple percentage rating were used for analysing the data while simple regression technique were used for testing the hypotheses. It was revealed, among other things that staff development and training have significant positive effect on workers’ efficiency; that provision of working tools and equipment have significant positive effect on patients’ satisfaction; and that issuance of honorary awards and recognition have significant positive effect on worker’s attitude. In its recommendation, the study submitted that since teaching hospitals double as health and educational institutions, they should provide professional development programmes for in-house staff. Furthermore, since they are public service institutions, the government should do well to provide modern innovative working tools to help staff members update their on-the-job skills and boost patients’ confidence on the quality of healthcare services rendered in these institutions.

Oversampling techniques for imbalanced data in regression

Our study addresses the challenge of imbalanced regression data in Machine Learning (ML) by introducing tailored methods for different data structures. We adapt K-Nearest Neighbor Oversampling-Regression (KNNOR-Reg), originally for imbalanced classification, to address imbalanced regression in low population datasets, evolving to KNNOR-Deep Regression (KNNOR-DeepReg) for high-population datasets. For tabular data, we also present the Auto-Inflater neural network, utilizing an exponential loss function for Autoencoders. For image datasets, we employ Multi-Level Autoencoders, consisting of Convolutional and Fully Connected Autoencoders. For such high-dimension data our approach outperforms the Synthetic Minority Oversampling Technique for Regression (SMOTER) algorithm for the IMDB-WIKI and AgeDB image datasets. For tabular data we conducted a comprehensive experiment using various models trained on both augmented and non-augmented datasets, followed by performance comparisons on test data. The outcomes revealed a positive impact of data augmentation, with a success rate of 83.75% for Light Gradient Boosting Method (LightGBM) and 71.57% for the 18 other regressors employed in the study. This success rate is determined by the frequency of instances where models performed better when augmented data was used compared to instances with no augmentation. Access to the comparative code can be found in GitHub.

Predicting Crop Yields in Indian Agriculture Using Machine Learning

Agriculture provides the backbone of India's economy both within and between. The paper below provides a brief insight into the creation of a fresh model is currently in progress to estimate the variability in the crop yields across various regions in India. Via this simple setting-based approach, such parameters as year, district, season and area can be used to predict the amounts of crop outputs for certain years. By employing sophisticated regression techniques, Kernel Ridge, Lasso, and ENet algorithms are used, to improve the yield prediction precision. Additionally, it implies a technique of Scaled Regression which is specifically designed to improve the accuracy of the data resulting as long-term predictions. Key terms: Crop yield forecasting, Lasso algorithm, Kernel Ridge algorithm, ENet algorithm, Stacked Regression.

Exploring the Role of Cognitive and Physical Limitations in Explaining Identity Theft Victimization

ABSTRACT We use data from the 2018 Identity Theft Supplement of the U.S. National Crime Victimization Survey to examine the relationship between cognitive or physical limitations and identity theft victimization from a Routine Activities theory perspective. Using the binary logistic regression and multiple imputation technique for handling the missing data, we regressed several outcomes of identity theft victimization on cognitive or physical limitations while controlling for online and offline routines including protective strategies, individual characteristics, and previous victimization. We also tested whether age moderates the relationship between cognitive or physical limitations and victimization outcomes. The results suggest that the relationship between cognitive or physical limitations and victimization is generally significant and positive except for credit card victimization Our findings also revealed that some of these relationships are moderated by age.

Machine Learning-based Extrapolation of Crop Cultivation Cost

It is important to comprehend the relation between operational expenses such as labour, seed, irrigation, insecticides, fertilizers and manure costs necessary for the cultivation of crops. A precise cost for the cultivation of crops can offer vital information for agricultural decision-making. The main goal of the study is to compare machine learning (ML) techniques to measure relationships among operational cost characteristics for predicting crop cultivation costs before the start of the growing season using the dataset made available by the Ministry of Agriculture and Farmer Welfare of the Government of India. This paper describes various ML regression techniques, compares various learning algorithms as well as determines the most efficient regression algorithms based on the data set, the number of samples and attributes. The data set used for predicting the cost with 1680 instances includes varying costs for 14 different crops for 12 years (2010-2011 to 2021-2022). Ten different ML algorithms are considered and the crop cultivation cost is predicted. The evaluation results show that Random Forest (RF), Decision Tree (DT), Extended gradient boosting (XR) and K-Neighbours (KN) regression provide better performance in terms of coefficient of determination (R2), root mean square error (RMSE) and mean absolute error (MAE) rate while training and testing time. This study also compares different ML techniques and showed significant differences using the statistical analysis of variance (ANOVA) test. The optimal hyperparameters for the ML models are found using the gridsearchCV and randomizedsearchCV functions, which improves the model's capacity for generalisation.

Enhancing Transparency of Climate Efforts: MITICA’s Integrated Approach to Greenhouse Gas Mitigation

Under the Paris Agreement, countries must articulate their most ambitious mitigation targets in their Nationally Determined Contributions (NDCs) every five years and regularly submit interconnected information on greenhouse gas (GHG) aspects, including national GHG inventories, NDC progress tracking, mitigation policies and measures (PAMs), and GHG projections in various mitigation scenarios. Research highlights significant gaps in the definition of mitigation targets and the reporting on GHG-related elements, such as inconsistencies between national GHG inventories, projections, and mitigation targets, a disconnect between PAMs and mitigation scenarios, as well as varied methodological approaches across sectors. To address these challenges, the Mitigation-Inventory Tool for Integrated Climate Action (MITICA) provides a methodological framework that links national GHG inventories, PAMs and GHG projections, applying a hybrid decomposition approach that integrates machine learning regression techniques with classical forecasting methods for developing GHG emission projections. MITICA enables mitigation scenario generation until 2050, incorporating over 60 PAMs across Intergovernmental Panel on Climate Change (IPCC) sectors. It is the first modelling approach that ensures consistency between reporting elements, aligning NDC progress tracking and target setting with IPCC best practices while linking climate change with sustainable economic development. MITICA’s results include projections that align with observed trends, validated through cross-validation against test data, and employ robust methods for evaluating PAMs, thereby establishing its reliability.

Characterizing diseases using genetic and clinical variables: A data analytics approach

AbstractPredictive analytics is crucial in precision medicine for personalized patient care. To aid in precision medicine, this study identifies a subset of genetic and clinical variables that can serve as predictors for classifying diseased tissues/disease types. To achieve this, experiments were performed on diseased tissues obtained from the L1000 dataset to assess differences in the functionality and predictive capabilities of genetic and clinical variables. In this study, the k‐means technique was used for clustering the diseased tissue types, and the multinomial logistic regression (MLR) technique was applied for classifying the diseased tissue types. Dimensionality reduction techniques including principal component analysis and Boruta are used extensively to reduce the dimensionality of genetic and clinical variables. The results showed that landmark genes performed slightly better in clustering diseased tissue types compared to any random set of 978 non‐landmark genes, and the difference is statistically significant. Furthermore, it was evident that both clinical and genetic variables were important in predicting the diseased tissue types. The top three clinical predictors for predicting diseased tissue types were identified as morphology, gender, and age of diagnosis. Additionally, this study explored the possibility of using the latent representations of the clusters of landmark and non‐landmark genes as predictors for an MLR classifier. The classification models built using MLR revealed that landmark genes can serve as a subset of genetic variables and/or as a proxy for clinical variables. This study concludes that combining predictive analytics with dimensionality reduction effectively identifies key predictors in precision medicine, enhancing diagnostic accuracy.

Improving Anticancer Drug Selection and Prioritization via Neural Learning to Rank.

Personalized cancer treatment requires a thorough understanding of complex interactions between drugs and cancer cell lines in varying genetic and molecular contexts. To address this, high-throughput screening has been used to generate large-scale drug response data, facilitating data-driven computational models. Such models can capture complex drug-cell line interactions across various contexts in a fully data-driven manner. However, accurately prioritizing the most effective drugs for each cell line still remains a significant challenge. To address this, we developed multiple neural ranking approaches that leverage large-scale drug response data across multiple cell lines from diverse cancer types. Unlike existing approaches that primarily utilize regression and classification techniques for drug response prediction, we formulated the objective of drug selection and prioritization as a drug ranking problem. In this work, we proposed multiple pairwise and listwise neural ranking methods that learn latent representations of drugs and cell lines and then use those representations to score drugs in each cell line via a learnable scoring function. Specifically, we developed neural pairwise and listwise ranking methods, Pair-PushC and List-One on top of the existing methods, pLETORg and ListNet, respectively. Additionally, we proposed a novel listwise ranking method, List-All, that focuses on all the effective drugs instead of the top effective drug, unlike List-One. We also provide an exhaustive empirical evaluation with state-of-the-art regression and ranking baselines on large-scale data sets across multiple experimental settings. Our results demonstrate that our proposed ranking methods mostly outperform the best baselines with significant improvements of as much as 25.6% in terms of selecting truly effective drugs within the top 20 predicted drugs (i.e., hit@20) across 50% test cell lines. Furthermore, our analyses suggest that the learned latent spaces from our proposed methods demonstrate informative clustering structures and capture relevant underlying biological features. Moreover, our comprehensive evaluation provides a thorough and objective comparison of the performance of different methods (including our proposed ones).